← All models

GLM-Z1-9B-0414

Published Publisher Zhipu AI

Model Positioning

GLM-Z1-9B-0414 is a small reasoning model with approximately 9 billion parameters released by Zhipu AI, intended for resource-constrained environments and lightweight deployment. Zhipu AI lists it as a reasoning model and makes its model weights available; the official model page specifies an MIT license. (github.com)

Capabilities and Uses

  • According to the official description, it excels at mathematical reasoning and general tasks and can be used for related text reasoning and question-answering scenarios.
  • The model page lists Chinese, English, and text generation; it can be deployed and used with tools such as Transformers and vLLM. (github.com)

Context and Limitations

  • The officially listed sequence length is 32K → 128K: the model was trained with a native 32K context, and enabling YaRN extension is recommended when the combined input and output length exceeds 32K; Zhipu AI also recommends considering YaRN for the GLM-Z1 series when input exceeds 8,192 tokens. Extended-context performance depends on the actual configuration. (github.com)
  • Model outputs may contain errors, and important conclusions should be verified by humans or against external sources. The resources required and results achieved for a specific deployment depend on the inference framework and configuration.

Pricing

As of 2026-11-18, a review of Zhipu AI's official domestic pricing documentation and Z.ai's official international pricing documentation did not verify an official API input or output price for this specific model version; prices for other GLM models cannot be substituted. The domestic platform's pricing page requires JavaScript, so its full contents cannot be read directly. Self-deployment may also incur compute and operations costs. (docs.bigmodel.cn)

Release and Official Sources